key takeaways
If you only read 30 seconds of this article:
- Industry estimates suggest manufacturers carry 20-30% excess inventory while facing 10-15% stockout risk, the result of optimizing each site in isolation.
- Centralize criticality ranking across sites so high-consequence parts (those that stop production lines) pool inventory, while low-consequence parts reduce to zero.
- Standardize stocking policies across locations with the same assets; a bearing that fails twice per year should have the same safety-stock rule on every rig, not different rules per site.
- Measure success by two metrics: days of working capital recovered and uptime improvement, not inventory turns, a metric that penalizes holding critical spares.
How much duplicate safety stock is in your network?
A quick call maps where pooling would release working capital without adding risk.
Short answer: Industry estimates suggest the average asset-intensive manufacturer carries 20-30% excess MRO inventory and simultaneously faces stockout risk on 10-15% of critical parts, consistent with Verusen's experience across hundreds of implementations. This paradox exists because single-site optimization tools cannot coordinate stocking decisions across multiple ERPs and locations, so each plant overbuys what it can see and under-stocks what it cannot. A multi-site MRO inventory strategy connects inventory visibility, centralized criticality ranking, and cross-site pooling into one coordinated framework that eliminates both excess and shortage simultaneously.
MRO inventory strategy: An MRO inventory strategy is a coordinated decision framework that determines what to stock, how much, and where across multiple sites and ERPs based on asset criticality, demand history, and supply risk, not forecast accuracy or rotation speed. Unlike finished-goods inventory, MRO spare parts fail on an unpredictable schedule, so the strategy must rank by criticality and consequence, not by sales velocity.
What Is Multi-Site MRO Inventory Strategy (and Why Single-Site Thinking Fails)
Multi-site MRO inventory strategy is a framework that aligns stocking decisions, criticality assignments, and reorder points across all plants in your network at the same time, rather than letting each site manager optimize their own maintenance, repair and operations inventory in isolation. Industry estimates suggest the average asset-intensive manufacturer carries 20-30% excess MRO inventory and simultaneously faces stockout risk on 10-15% of critical parts, consistent with Verusen's experience across hundreds of implementations. This contradiction exists because single-site optimization treats each plant as an island, duplicating safety stock across locations while missing aggregate demand signals that only appear when you sum failure patterns across all sites.
Why Single-Site Optimization Fails Across a Network
A bearing fails twice in five years at Plant A. Standard safety stock formulas require demand history; two data points returns zero recommended stock. Plant A orders zero and the line stops for three weeks. Meanwhile, Plant B carries the same bearing but records no failures in five years, so the formula recommends zero there too. Both plants end up understocked on a critical part because neither site has enough failure history to trigger the formula. Each plant manager, working alone, cannot see the aggregate pattern: the bearing fails four times across your network every five years, a signal buried under site-level noise.
Single-site thinking also duplicates safety stock. If each of your 10 plants manages its own bearing inventory independently, you may carry 10 safety stocks when network strategy would hold 2 or 3 at centralized hubs and distribute spares on demand. Georgia Pacific, operating 110 US sites across multiple ERP systems, discovered this exact problem: hundreds of people making independent stocking decisions meant $55M in inventory was identified as excess, and 2,900 materials were flagged at stockout risk simultaneously, the signature of a fragmented network, not an optimized one.
The Multi-Site Framework: Three Steps to Network Alignment
- Aggregate failure history across every site: sum failure events for each material network-wide so the true failure rate emerges from what looks like noise at any single location.
- Classify by criticality and lead time, not demand: rank each material by production consequence (line-stop, degrade, workaround) combined with supplier lead time, using the network-wide failure rate from step 1.
- Apply a placement rule per material: decide hub, spoke, or both using the placement matrix below — stock critical short-MTBF parts near the assets, consolidate everything else centrally.
A major global mining company with 17 sites across three ERP systems applied this framework and identified $96.8M in excess inventory during the evaluation phase. The key outcome: centralized decisioning replaced dozens of independent stocking policies with one network logic, enabling the maintenance team to flag critical stockouts before they happened rather than respond to line stops after the fact.
Why Single-Site MRO Optimization Creates Multi-Site Stockouts
Each plant in a multi-site manufacturer optimizes its own MRO inventory in isolation, which works for finished goods but breaks catastrophically for spare parts. A bearing that fails once every two years at Plant A and once every three years at Plant B looks like low-priority scrap when each site analyzes its own demand history separately, but is actually a shared critical risk that single-site optimization cannot see.
Standard safety stock formulas require demand history to calculate reorder points. A bearing that fails twice in five years has no meaningful history, so the formula returns zero. Plant A orders zero. Plant B, by chance, has two recent failures and orders safety stock. Then the bearing fails at Plant A, the line stops for three weeks, and the part that looked like inventory bloat at the corporate level was actually the difference between running and stopped.
A Fortune 500 CPG manufacturer with 41 sites across multiple regions faced exactly this problem: each plant managed its own stocking policies independently, creating pockets of dead stock at some locations and critical shortages at others. When they optimized inventory across all 41 sites simultaneously rather than site-by-site, they identified $63M in excess inventory and verified $60M in actionable savings, based on Verusen customer results. The shift from plant-level to network-level visibility also cut material review time from over 20 minutes to 4 minutes per decision, because the system could instantly flag whether a part was overstocked anywhere in the network.
Why site-by-site optimization hides the real problem
Single-site optimization assumes each plant's demand is independent. For production equipment that rarely fails, demand IS sparse and local. But across a network of identical or similar assets, the same part fails somewhere every few months, even if it fails only once every two years at any single location. When each site optimizes in isolation, it sees only its own scarcity and orders defensively. The network sees abundance and obsolescence.
The problem deepens when plants carry different ERP systems or when historical data is incomplete. Plant A might have three years of maintenance records; Plant B, six months. Plant A's formula says "no demand signal," so it stocks zero. Plant B has slightly more history and stocks some. Neither plant knows that a centralized view would show the part fails regularly across the network, just not predictably at any single site. This is why industry estimates suggest the average asset-intensive manufacturer carries 20-30% excess MRO inventory while simultaneously facing stockout risk on 10-15% of critical parts, consistent with Verusen's experience across hundreds of implementations. The contradiction resolves only when you optimize across the entire network at once.
How to Build an MRO Inventory Strategy Across Multiple Plants
A multi-site MRO inventory strategy pools failure data across all locations, then places stock where it minimizes total downtime risk and working capital, not where each plant happened to fail last. Single-site optimization creates a false choice: stock at every location and carry 20-30% excess inventory, or centralize everything and risk stockouts on critical parts — industry estimates suggest the average asset-intensive manufacturer faces both simultaneously, consistent with Verusen's experience across hundreds of implementations.
The solution is network-aware placement. You aggregate failure history across all plants to reveal the true rate of each failure (not the noise of a single location), classify materials by criticality and lead time, then apply a placement rule for each material: stock at hub, at multiple spokes, or both. Georgia Pacific executed this logic across 110 US sites running 4 ERP systems, identifying $55M in savings and reducing material review time from hours per decision to minutes by centralizing placement decisions to one team of 7, replacing hundreds of independent plant choices (based on Verusen customer results).
Stocking Placement Decision Matrix
| Lead Time vs. MTBF | Criticality Tier | Stocking Location | Why |
|---|---|---|---|
| Lead time < MTBF | High (stops production) | Hub + nearest 2-3 spokes | Fast enough to stock locally; criticality justifies redundancy |
| Lead time < MTBF | Medium (slows work) | Hub only | Fast replenishment sufficient; don't stock every location |
| Lead time > MTBF | High (stops production) | Hub + regional spokes | Lead time too long for emergency reorder; pre-position regionally |
| Lead time > MTBF | Medium or Low | Hub only or consignment | Justify central stock only if failure rate warrants it across network |
Concrete example: a motor with 12-week lead time and failures across 4 plants stages at the central warehouse because the lead time exceeds the average time between failures, no single plant can reorder fast enough. A bearing with 2-week lead time and high criticality stages at both the hub and the two nearest plants, so a critical failure can be filled within hours instead of waiting for a warehouse shipment. Learn more on how to determine safety stock for spare parts using your network-wide failure history.
This works because most ERP systems track inventory locally, so rare failures appear rare everywhere. Aggregating across all plants reveals the true failure rate and eliminates both stockouts on critical parts and excess stock on low-movers. The payoff is real: Seadrill optimized MRO across 17 rigs on Maximo by applying hub-and-spoke logic, verifying $3.3M in working capital savings in phase 1 alone (based on Verusen customer results).
See network-wide failure rates on your parts
We aggregate failure history across your ERPs so rare events become plannable.
Why Criticality, Not Historical Demand, Drives Multi-Site Stocking Decisions
Spare parts fail on criticality, not on a predictable schedule, and for multi-site manufacturers, this distinction is the difference between a stocking policy that prevents production loss and one that ties up millions in dead stock while you're short on the parts that matter. Standard demand-planning models such as SAP IBP and Maximo return zero stock recommendations for parts with sparse failure history because they were built for finished goods, not for maintenance inventory that doesn't follow a sales forecast.
In multi-site networks, each plant has its own equipment footprint, maintenance intervals, and run patterns. A bearing that fails quarterly at Plant A may never fail at Plant B because the asset is newer or runs a different duty cycle. Demand-history formulas cannot distinguish between these scenarios. A major US energy company reviewed 45,000 materials using Maximo and had to manually classify them by impact before optimization could begin, based on Verusen customer results.
The Criticality Decision Framework: Three Questions to Replace Demand History
- Does production stop if this part fails? Parts that halt a line are critical. Parts that degrade performance or quality but allow continued operation are important. Parts with substitutes or workarounds are standard.
- What is the lead time and replacement downtime combined? A part with a 6-week lead time and 6-hour replacement downtime is critical even if it fails once per decade. A part with a 2-day lead time and 4-hour replacement downtime is less critical, even if it fails monthly.
- What is the mean time between failures at each site? Network-wide MTBF replaces demand history. A bearing that fails every 90 days at Plant A and never at Plant B triggers stocking at Plant A only, not both.
How to Build a Multi-Site Stocking Allocation Matrix
Map each part type to a criticality tier, then apply a stocking location rule based on lead time and failure distribution across your network. A Fortune 500 CPG manufacturer across 41 sites reduced material review time from over 20 minutes to 4 minutes by classifying all 6,200+ MRO items into three tiers and applying location rules, based on Verusen customer results.
| Part Type | Criticality | Lead Time | Stocking Location Rule | Safety Stock Formula |
|---|---|---|---|---|
| Motor (production line drive) | Critical | 12 weeks | Central hub warehouse: 2 units. Nearest 2 plants: 1 unit each. | MTBF + 2 sigma buffer |
| Bearing (reduces output if missing) | Important | 2 weeks | Central hub: 1 unit. All plants with that asset: 1 unit each. | MTBF + 1 sigma buffer |
| Coupling (has 2-day substitute) | Standard | 3 days | Central hub only, order-to-stock. | MTBF + 0 sigma (emergency stock) |
Seadrill, a global offshore operator with 17 rigs and Maximo, implemented this hub-and-spoke model after Verusen identified $48M in MRO inventory, based on Verusen customer results. High-criticality items like subsea control modules staged at a central shorebase; medium-criticality spares distributed to 2 to 3 nearest rigs; low-criticality items remained order-to-stock. The result: enabled rapid failure response without carrying duplicates across all 17 assets.
Go deeper: this article supports our pillar guide, MRO Inventory Optimization: The Complete Guide. Related: MRO management strategies: five rules for spare parts inventory.
Further reading: MRO spares inventory optimization guide, spare parts inventory management guide, and MRO inventory optimization best practices.
Build the hub-and-spoke plan
Talk through a placement matrix for your sites with an MRO expert.
Hub-and-spoke placement without AI — the default in Maximo- or SAP-planned networks — still computes each site's safety stock from local demand history; the pooling logic above only works when failure data is aggregated network-wide first, which is precisely the step those systems skip.
Frequently asked questions
Single-site strategy optimizes inventory for one location's failure patterns and lead times; multi-site strategy must balance centralized purchasing power against distributed demand variability and regional supply-chain risk. A single plant can stock to its own failure history, but a 41-site manufacturer faces the problem that a bearing failing twice across the entire network in five years leaves each site with no history to forecast from. Multi-site manufacturers must implement hub-and-spoke stocking policies, cross-site sharing protocols, and centralized visibility into which parts are critical at which locations to avoid both excess inventory in low-risk zones and stockout risk in high-risk ones.
Map failure criticality and lead time by part and by location, then set safety stock only on parts where stockout cost exceeds holding cost. Industry estimates suggest the average asset-intensive manufacturer faces stockout risk on 10 to 15% of critical parts while simultaneously carrying 20 to 30% excess inventory, consistent with Verusen's experience across hundreds of implementations. A Fortune 500 CPG manufacturer reduced material review time from over 20 minutes to 4 minutes by centralizing which parts needed higher safety stock across 41 sites. Implement a centralized decision layer that flags high-criticality, long-lead materials and stages them in regional hubs, then use demand-sensing on the remaining parts to flow inventory just-in-time.
Each site orders independently to its own perceived risk, creating duplicate safety stock across locations for the same part, while no location is accountable for inventory it doesn't control. A leading gold mining company identified $96.8M in excess MRO inventory across 17 sites running three separate ERPs, based on Verusen customer results. When one plant fears a stockout and orders extra, and another plant does the same, the system carries redundant stock nobody claims ownership of. Centralized inventory visibility, showing which parts are already staged at which locations and their lead times, eliminates the duplicative safety-stock ordering that drives the 20 to 30% overage.
Classify parts first by criticality and lead time across the entire network, then decide which locations should hold reserve stock and which should rely on rapid redistribution from central hubs based on each part's failure cost at each site. This sequencing ensures you're not duplicating expensive inventory at low-risk locations just because one site fears a stockout. Standard formulas require demand history; for a part failing twice in five years, the formula returns zero, forcing blind orders or stockout risk. A Fortune 500 beverage producer with 130+ plants across 6 global zones identified $55M in unnecessary safety stock by modeling which parts genuinely needed elevated reserve levels at each location and consolidating the rest, based on Verusen customer results.
Yes, connect directly to your existing ERP, EAM, or P2P system and run the analysis on live data without any data cleanse or system outage. Verusen optimizes inventory across multiple sites and returns ROI in weeks without requiring a data cleanup first, based on Verusen customer results. A Fortune 500 CPG manufacturer identified $63M and verified $60M in MRO inventory savings across 41 sites while keeping all plants running on their original schedules. The optimization identifies which parts you can safely reduce, which need higher safety stock, and which can be consolidated into regional hubs, all without touching your production schedule.
PN
- Paul Noble
- Founder & CEO, Verusen
Paul founded Verusen to bring AI-native systems of record to industrial materials. He has spent 15+ years working alongside F&B, oil & gas, and manufacturing operators on the MRO data problem.
